On Thursday, November 2, 2017 at 5:09:14 PM UTC-4, Ian Denhardt wrote: > > Quoting Krzysztof Sakrejda (2017-11-02 05:36:56) > > > Thanks, I guess that does mean it'll work as long as the floating > point > > implementation is the same. > > > Basically everything uses IEEE-754 these days,
This I am aware of but this is working with statistical software (mc-stan.org) that regularly runs on clusters of varying vintage and we have plenty of users who don't talk to us so I wouldn't be surprised if it became an issue. Recently somebody on the users list was asking for installation help on a cluster with no internet connection so... ¯\_(ツ)_/¯ > so the representation of > floating point values isn't really any less portable than it is for > signed integers (where theoretically the standards allow for one's > compliment and signed-magnitude, but you'll be hard pressed to find > anything but two's compliment): > > > https://en.wikipedia.org/wiki/Floating-point_arithmetic#IEEE_754:_floating_point_in_modern_computers > > > You're welcome. > Thanks! Krzysztof > > -Ian > -- You received this message because you are subscribed to the Google Groups "Cap'n Proto" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. Visit this group at https://groups.google.com/group/capnproto.
